Chromatin loops,illegitimate recombination,and genome evolution

Chromosomal rearrangements frequently occur at specific places (“hot spots”) in the genome. These recombination hot spots are usually separated by 50–100 kb regions of DNA that are rarely involved in rearrangements. It is quite likely that there is a correlation between the above‐mentioned distances and the average size of DNA loops fixed at the nuclear matrix. Recent studies have demonstrated that DNA loop anchorage regions can be fairly long and can harbor DNA recombination hot spots. We previously proposed that chromosomal DNA loops may constitute the basic units of genome organization in higher eukaryotes. In this review, we consider recombination between DNA loop anchorage regions as a possible source of genome evolution.     

Regulation of the operon encoding ribonucleotide reductase: role of the negative sites in nrd repression.

Expression of the nrd genes was previously shown to be controlled by both positive and negative regulation (C. K. Tuggle and J. A. Fuchs, EMBO J. 5:1077-1085, 1986). Two regions, one located 5' and one located 3' of the nrd promoter (nrdP), were identified as negative regulatory sites since deletion of these sequences increased nrd expression. These regions of DNA have sequence similarities, and a looping mechanism was proposed to explain the requirement for two distinct sites in nrd repression. To investigate the role of these sequences in regulating nrd, a gel electrophoresis assay was used to detect the proteins that bind to the nrd regulatory sites. A protein that bound to restriction fragments containing the negative regulatory sites but not to other DNA fragments was identified in cell extracts and was partially purified. DNase I footprinting experiments showed that the binding protein protects the 5' negative site previously identified in vivo. The 3' negative site also identified in vivo was not required in vitro for high-affinity protein binding to the 5' site, but lower-affinity binding to this site could be detected. Specific binding to the 5' site was found to be elevated approximately 10-fold in crude extracts from thymine-starved cells as compared with that in extracts from unstarved cells. This higher activity was also evident in purified preparations, suggesting that thymine starvation increases the expression of the negative regulatory protein. The finding that a purified protein preparation binds both negative regulatory sites indicates that this preparation contains the nrd repressor protein or proteins. Insertion of 37 base pairs (3.5 helix turns) of DNA at a HpaII site or 35 base pairs (3.3 turns) at a MnlI site between the 5' regulatory sites and nrdP abolished the increase in nrd expression resulting from thymine starvation in vivo, but negative regulation appeared to be less affected than when either negative site was deleted. Insertion of DNA in these constructs was shown not to affect repressor binding in vitro, indicating either that a simple model of DNA looping to bring equivalent operator sites into physical proximity does not explain repression at nrd or that the distance between sites is sufficient that helical turns are of little importance.     

Collaborative competition mechanism for gene activation in vivo

The mechanism by which gene regulatory proteins gain access to their DNA target sites is not known. In vitro, binding is inherently cooperative between arbitrary DNA binding proteins whose target sites are located within the same nucleosome. We refer to such competition-based cooperativity as collaborative competition. Here we show that arbitrarily chosen foreign DNA binding proteins, LexA and Tet repressor, cooperate with an adjacently binding endogenous activator protein, Gcn4, to coactivate expression of chromosomal reporter genes in Saccharomyces cerevisiae. Coactivation requires that the cooperating target sites be within a nucleosome-length distance; it leads to increased occupancy by Gcn4 at its binding site; and it requires both Gcn5 and Swi/Snf which, at an endogenous Gcn4-dependent promoter, act subsequent to Gcn4 binding. These results imply that collaborative competition contributes to gene regulation in vivo. They further imply that, even in the presence of the cell's full wild-type complement of chromatin remodeling factors, competition of regulatory proteins with histone octamer for access to regulatory target sites remains a quantitative determinant of gene expression levels. We speculate that initial target site recognition and binding may occur via spontaneous nucleosomal site exposure, with remodeling factor action required downstream to lock in higher levels of regulatory protein occupancy.     

DNA tape measurements of AraC

A new method for measuring distances between points in the AraC-DNA complex was developed and applied. It utilizes variable lengths of single-stranded DNA that connect double-stranded regions containing the two half-site binding sequences of AraC. These distances plus the protein interdomain linker distances are compatible with two classes of structure for the dimeric AraC gene regulatory protein. In one class, the N-terminal regulatory arm of one dimerization domain is capable of interacting with the DNA-binding domain on the same polypeptide chain for a cis interaction. In the other class, the possible arm-DNA-binding domain interaction is trans, where it adds to the dimerization interface.     

Examining the contribution of a dA+dT element to the conformation of Escherichia coli integration host factor-DNA complexes.

DNA binding proteins that induce structural changes in DNA are common in both prokaryotes and eukaryotes. Integration host factor (IHF) is a multi-functional DNA binding and bending protein of Escherichia coli that can mediate protein-protein and protein-DNA interactions by bending DNA. Previously we have shown that the presence of a dA+dT element 5'-proximal to an IHF consensus sequence can affect the binding of IHF to a particular site. In this study the contribution of various sequence elements to the formation of IHF-DNA complexes was examined. We show that IHF bends DNA more when it binds to a site containing a dA+dT element upstream of its core consensus element than to a site lacking a dA+dT element. We demonstrate that IHF can be specifically crosslinked to DNA with binding sites either containing or lacking this dA+dT element. These results indicate the importance of flanking DNA and a dA+dT element in the binding and bending of a site by IHF.